The Problem
Modern large scale radio telescopes are made of multiple antennas whose signal needs to be combined to form a clear picture of the night sky. During this process, the antenna system needs to be calibrated to digitally remove noise source and imperfection in the telescope setup. Classical calibration techniques are powerful but computationally demanding.
The Quantum Solution
Together with Astron, we have implemented quantum linear solvers relying on two different backends namely IBM quantum and DWave. We have integrated these quantum solvers in radioastronomy pipeline developed for the operation of the HERA telescope and benchmark their performance against classical linear solvers. We have demonstrated that the performance of these solvers, in particular the ones relying on quantum annealers can quickly provide a first approximation of the solution that further be refined using classical approaches. The solvers are freely available on the QAL github an can readily be reused for any application requiring the solution of linear solvers.
The Benefits
The application of quantum computing methods for solving linear system of equations has always a been a promising avenue for quantum computing. The integration of these solvers in complete processing pipelines shows the possibility to use these solvers and offer a path forward to integrate quantum computers in the design and operation of large scale data acquisition pipeline of radioastronomy.
Links
Link to the paper : https://arxiv.org/abs/2310.11932
Link to the code : https://github.com/QuantumRadioAstronomy/hera_cal_quantum
Link to the VQLS library : https://github.com/QuantumApplicationLab/vqls-prototype
